Komal Bankim Shah MD, Presenter: Nothing to Disclose

Dawid Schellingerhout MD, Abstract Co-Author: Speakers Bureau, ArthroCare Corporation

Jeffrey S. Weinberg MD, Abstract Co-Author: Nothing to Disclose

Ganesh Rao MD, Abstract Co-Author: Nothing to Disclose

Sujit Prabhu MD, Abstract Co-Author: Nothing to Disclose

Ashok J. Kumar MD, Abstract Co-Author: Nothing to Disclose

Intra-operative imaging allows postoperative scanning in the operating room. We compared final intraoperative diffusion weighted imaging (DWI) sequences with DWI obtained 24 hours after surgery (standard imaging) to evaluate the optimal time-point for detection of perioperative ischemia.

We reviewed 55 patients undergoing intraoperative MRI during resection of suspected glioblastoma between 10/2006 and 2/2009. We compared the final intra-operative DWI and postoperative DWI in each case.

Of 55 patients, 53 had glioblastoma, one had gliosarcoma and one had treatment related necrosis. At final intraoperative scan, 25 of 55 (45%) had some DWI findings. At the 24h postoperative scan (4-96 hours between scans, average 26 hours), 13 additional patients became DWI positive, for a total of 38 of 55 (69%). No patient became DWI negative. 11 patients with DWI findings on final intraoperative scan progressed, resulting in 24 of 55 patients with new or progressive DWI findings (44%). Fourteen patients with positive DWI findings that remained stable had rim-like marginal restricted diffusion only. The 11 patients with progressive DWI findings showed 2 new infarcts, 1 infarct enlargement, and 1 hematoma, while 7 had progression of marginal restricted diffusion. The 13 patients who converted from negative to positive DWI findings showed 1 new infarct and 3 hematomas, while 9 developed marginal restricted diffusion. Twenty-one of 24 patients with new or progressive diffusion restriction had follow-up imaging available for review. In 10 of 21 patients, marginal restricted diffusion progressed to nodular Gadolinium-enhancement on follow-up MRI, mimicking tumor recurrence in 6 of 10 patients.

The 24h scan showed additional information in 24 of 55 patients (44%). In 10 of 55 patients (18%) the information was clinically important (infarcts and tumor mimics).

Even if intraoperative MRI is available, we suggest that patients still undergo routine scanning at 24h with DWI.

Shah, K, Schellingerhout, D, Weinberg, J, Rao, G, Prabhu, S, Kumar, A, Postoperative Diffusion Abnormalities in the Follow-up of Glioblastoma: Final Intraoperative MRI versus 24-hour Postoperative Study.  Radiological Society of North America 2009 Scientific Assembly and Annual Meeting, November 29 - December 4, 2009 ,Chicago IL. http://archive.rsna.org/2009/8009258.html